BigCatc/ultrafeedback-cleaned-rm-annot

该数据集包含多个字段，其中包括两个字符串类型的字段rejected和chosen，可能代表被拒绝和被选择的项目或选项。prompt字段可能表示相关的提示信息。另外还有四个浮点数字段，分别为chosen_finegrained、rejected_finegrained、chosen_score和rejected_score，这些可能代表更细致的评分或分类结果。数据集分为训练集，包含222560个示例，总大小为641937063字节。具体的应用场景和详细描述在README中未提及。

The dataset includes multiple fields, among which there are two string-type fields, rejected and chosen, which may represent rejected and selected items or options. The prompt field may represent related prompt information. There are also four floating-point fields: chosen_finegrained, rejected_finegrained, chosen_score, and rejected_score, which may represent more detailed scoring or classification results. The dataset is split into a training set, containing 222560 examples, with a total size of 641937063 bytes. The specific application scenario and detailed description are not mentioned in the README.